A semiparametric regression cure model for doubly censored data

Area of Science:

• Biostatistics
• Survival Analysis
• Medical Statistics

Background:

• Doubly censored failure time data involves observations on two event times, both subject to censoring.
• Existing statistical methods lack established approaches for analyzing such data when a cured subgroup is present.
• Acquired immune deficiency syndrome (AIDS) cohort studies are a typical example where this data censoring occurs.

Purpose of the Study:

• To develop and present a regression analysis method for doubly censored failure time data with a cured subgroup.
• To address the limitations of current statistical methods in handling cured populations within interval-censored data.
• To provide a statistically sound approach for analyzing complex survival data in medical research.

Main Methods:

• A sieve approximation maximum likelihood approach is proposed for regression analysis.
• The method is designed to handle the complexities of doubly censored data and the presence of a cured subgroup.
• Asymptotic properties of the resulting estimators are theoretically established.

Main Results:

• The developed sieve maximum likelihood method demonstrates effectiveness in regression analysis of doubly censored data with cured subgroups.
• Extensive simulation studies confirm the method's good performance in practical scenarios.
• The approach provides reliable estimators for survival data analysis.

Conclusions:

• The proposed sieve approximation maximum likelihood method is a viable and effective tool for analyzing doubly censored failure time data with cured subgroups.
• The method offers advancements in survival analysis, particularly for complex medical datasets like AIDS cohort studies.
• The study provides a valuable statistical framework with demonstrated practical applicability and theoretical support.